White House Requests Delay in UK Access to New AI Models from OpenAI and Anthropic

Key Intelligence

• ■The White House has asked leading AI developers, OpenAI and Anthropic, to withhold two of their new 'frontier' AI models from UK testers.
• ■This request comes as the US plans its own review of these advanced AI systems, citing security concerns.
• ■The decision prevents the UK's AI Safety Institute from gaining early, pre-release access to these models, potentially impacting their testing and safety research efforts.
• ■The move highlights the US's cautious approach to regulating and evaluating cutting-edge AI technologies before wider international sharing.